Health-related quality of life and survival outcomes in prostate cancer

(1)

80/2020 ISBN 978-951-51-6762-0 (PRINT)

ISBN 978-951-51-6763-7 (ONLINE) ISSN 2342-3161 (PRINT) ISSN 2342-317X (ONLINE)

http://ethesis.helsinki.fi HELSINKI 2020

SUSANNE BERGIUS HEALTH-RELATED QUALITY OF LIFE AND SURVIVAL OUTCOMES IN PROSTATE CANCER

dissertationesscholaedoctoralisadsanitateminvestigandam universitatishelsinkiensis

DEPARTMENT OF PUBLIC HEALTH FACULTY OF MEDICINE

DOCTORAL PROGRAMME IN POPULATION HEALTH UNIVERSITY OF HELSINKI

HEALTH-RELATED QUALITY OF LIFE AND SURVIVAL OUTCOMES IN PROSTATE CANCER

SUSANNE BERGIUS

(2)

HEALTH-RELATED QUALITY OF LIFE AND SURVIVAL OUTCOMES IN PROSTATE CANCER

Susanne Bergius

DOCTORAL DISSERTATION

To be presented for public discussion with the permission of the Faculty of Medicine of the University of Helsinki, in Auditorium 2, Haartman Institute,

on the 29th of January, 2021 at 12 o’clock.

Espoo 2020

(3)

Supervised by

Professor (emeritus) Harri Sintonen, PhD Department of Public Health

University of Helsinki

Professor (emeritus) Risto P. Roine, MD, PhD Department of Health and Social Management University of Eastern Finland

Reviewed by

Professor Matti Laato, MD, PhD Department of Surgery

University of Turku

Adjunct Professor Leena Forma, PhD

Faculty of Social Sciences, University of Helsinki Faculty of Social Sciences, Tampere University Opponent

Adjunct Professor Juha Laine, PhD University of Eastern Finland

The Faculty of Medicine uses the Urkund system (plagiarism recognition) to examine all doctoral dissertations

Dissertationes Scholae Doctoralis Ad Sanitatem Investigandam Universitatis Helsinkiensis Doctoral Programme in Population Health (DOCPOP)

ISSN 2342-3161 (print) ISSN 2342-317X (online)

ISBN 978-951-51-6762-0 (print) ISBN 978-951-51-6763-7 (online) Hansaprint, Helsinki 2020

(4)

To my family

”Minä kun kanssas kuuta katson sinun kanssas kuuta katson kultainen kuu luo

sydämiin siltaa”

- espoolainen tuutulaulu

(5)

Abstract

The aim of health care is to maximize health, and in practice, health must be produced in the context of scarce resources. In order to make wise resource allocations, health economic analyses are needed, often in the context of Health Technology Assessment (HTA), which is then used for informed decision-making. In addition to effectiveness, informed decision-making should also be based on other values, such as equity and equality. Health economic analyses concentrate on assessing the economic value of interventions, whereas full HTA also considers other aspects of decision-making.

In recent years, the concept of value-based health care (VBHC) has emerged on the side of traditional health economic analyses. They both try to answer questions concerning the value of interventions in health care - i.e., the relationship of outcomes and investments needed. The objective in VBHC is also to find such actions that can improve the cost-effectiveness of health care dynamically over time.

Health-related quality of life (HRQoL) is a patient-reported outcome measure which, in combination with survival data, can produce quality-adjusted life years (QALYs) gained needed as the outcome measure in cost-utility analyses (CUA). This thesis generates real-world data for the use of health economic analyses and compares outcomes of conventional treatment strategies in PC. According to the systematic literature review in this study, preference-based HRQoL data in PC patients are still scarce.

Prostate cancer (PC) is the second most common cancer among men, and its incidence rates have been rising especially in Western countries due to increased use of prostate-specific antigen (PSA) testing. However, due to improved diagnosis and advanced treatments, mortality rates are not increasing at the same rate, but in some countries, such as in Finland, there has even been a modest decline. As PC has become more prevalent, there is an increased burden to patients and society, demanding more understanding of real-world effectiveness and cost-effectiveness of interventions.

(6)

We measured the HRQoL of PC patients in different stages of the disease (Local, Locally Advanced and Metastatic) and in patients undergoing different treatments with two HRQoL instruments, namely the generic 15D and disease-specific EORTC QLQ-C30. HRQoL data were obtained from 1050 and clinical background data from 1024 patients. The mean age of the patients at baseline was 66.5 years, and most of the patients were in an early stage of the disease as only 59 (6%) of the patients were metastatic at the time of the diagnosis. Even though the mean 15D score of Local and Locally advanced patients did not differ from that of the age-standardized male population, there was a statistically significant difference on the dimensions of depression and distress among all patient groups, which indicates that there are psychological side effects from the awareness of cancer diagnosis. Out of the five functioning scales in the EORTC QLQ-C30 instrument, patients in the Local and Locally advanced groups scored the lowest in emotional function, which can indicate anxiety, worrying, irritation, and/or depression due to the awareness of the diagnosis.

The four major treatment strategies during the first year after diagnosis were active surveillance (n=226), radiation (n=280), radical surgery (n=299) and hormonal treatment (n=62). The mean follow-up time in the survival analysis was 77.7 months, and at the end of the follow-up, 84.4% of patients were alive. Median overall survival was 53.8 months (95% CI 44.5 – 63.2 months) in the hormonal group, and median survival for the other groups was not reached. Prostate cancer was a rare cause of death, especially in the active surveillance and surgery groups. The hormonal treatment group had the lowest HRQoL and survival among the studied treatment groups, and consequently, also experienced the least number of QALYs during the two-year follow-up.

Outcomes of the three other treatment groups were similar in terms of HRQoL and overall survival, and thus also regarding the number of QALYs experienced. Our study provided evidence that baseline HRQoL, measured by 15D score or certain 15D dimensions, has prognostic value in assessing overall as well as PC-specific survival.

As shown by the literature review in this study, the use of generic preference-based instruments suitable for calculating QALYs among PC patients is scarce. Therefore, a regression model (mapping model) was built to predict the generic 15D score from disease-specific EORTC QLQ-C30 data. The explanatory power of the EORTC QLQ-C30 mapping model to the 15D score was as high as 79%, which indicates that EORTC scales explained well the variance of the 15D scores.

(7)

Tiivistelmä

Terveydenhuollon tavoite on maksimoida terveyttä, joka käytännössä tapahtuu niukkojen resurssien vallitessa. Jotta resurssien kohdentaminen olisi viisasta ja informoitua, laajemman terveydenhuollon menetelmien arvioinnin (engl. Health Technology Assessment, HTA) yhteydessä tarvitaan myös taloudellista arviointia. Taloudellisessa arvioinnissa tarkastellaan intervention taloudellista arvoa, kun taas menetelmäarviointi (HTA) käsittää myös muita viisaaseen päätöksentekoon tarvittavia elementtejä kuten tasapuolisuuden ja oikeudenmukaisuuden.

Viime vuosina perinteisen taloudellisen arvioinnin rinnalle on syntynyt käsite arvoon perustuvasta terveydenhuollosta (value-based health care, VBHC). Molemmat pyrkivät vastaamaan kysymykseen terveydenhuollon interventioiden arvosta eli vaikuttavuuden ja tarvittavien investointien suhteesta.

VBHC:n tavoite on myös löytää sellaisia toimia, jotka voivat parantaa terveydenhuollon kustannusvaikuttavuutta.

Kun intervention aikaansaama, potilaan itse raportoima muutos terveyteen liittyvässä elämänlaadussa (health-related quality of life, HRQoL) yhdistetään tietoon elossaoloajan muutoksesta, voidaan laskea muutos laatupainotetuissa elinvuosissa (quality-adjusted life year, QALY), jota tarvitaan vaikuttavuusmittarina kustannusutiliteettianalyysissä (cost-utility analysis, CUA). Tämän väitöskirjatutkimuksen puitteissa tuotettiin tietoa eturauhassyöpäpotilaiden terveyteen liittyvästä elämänlaadusta terveydenhuollon arjen olosuhteissa ja verrattiin yleisimpien eturauhassyövän hoitolinjojen vaikuttavuutta. Tutkimuksessa tehdyn järjestelmällisen kirjallisuuskatsauksen perusteella preferenssipohjaisten elämänlaatumittareiden käyttö eturauhassyöpäpotilailla on harvinaista.

Eturauhassyöpä on toisiksi yleisin syöpä miehillä ja sen ilmaantuvuus on kasvanut länsimaissa etenkin lisääntyneen prostataspesifisen antigeenin (PSA) testauksen myötä. Kehittyneen diagnosoinnin ja tehokkaiden hoitojen ansioista eturauhassyöpäkuolleisuus ei ole lisääntynyt samassa suhteessa ilmaantuvuuden kanssa ja joissain maissa, kuten Suomessa, kuolleisuudessa on ollut pientä laskua. Koska eturauhassyöpä on tullut yhä yleisemmäksi, se muodostaa kasvavan tautitaakan potilaille ja yhteiskunnalle, mikä vaatii parempaa ymmärrystä hoitojen vaikuttavuudesta ja kustannusvaikuttavuudesta terveydenhuollon arjessa.

(8)

Tässä väitöskirjatutkimuksessa mitattiin eturauhassyöpäpotilaiden terveyteen liittyvää elämänlaatua taudin eri vaiheissa (paikallinen, paikallisesti edennyt ja metastaattinen tauti) ja eri hoidoissa kahdella eri elämänlaatumittarilla, jotka olivat geneerinen 15D ja sairaus-spesifi EORTC QLQ-C30. Elämänlaatutieto saatiin 1050 potilaalta ja kliiniset taustatiedot 1024 potilaalta.

Potilaiden keski-ikä tutkimukseen tullessa oli 66,5 vuotta ja suurin osa heistä oli taudin varhaisessa vaiheessa. Metastaattisessa taudin vaiheessa oli tutkimuksen alussa 59 potilasta (6 %). Paikallisessa ja paikallisesti edenneessä taudissa 15D:n tuottaman keskimääräinen elämänlaatulukema ei eronnut ikävakioidun miesväestön lukemasta, mutta masennuksessa ja ahdistuneisuudessa havaittiin kaikissa taudin tiloissa tilastollisesti merkitsevä ero ikävakioituun miesväestöön verrattuna, mikä viittaa syöpädiagnoosin psyykkisiin vaikutuksiin. EORTC QLQ-C30:n viidestä toiminnallisesta ulottuvuudesta potilaat raportoivat alhaisimmat pisteet tunne-ulottuvuudessa, mikä voi viitata syöpädiagnoosiin liittyvään huoleen, pelkoon, hermostuneisuuteen ja/tai masennukseen.

Neljä merkittävintä hoitolinjaa ensimmäisen vuoden aikana diagnoosista olivat aktiivinen seuranta (n=226), sädehoito (n=280), radikaalileikkaus (n=299) ja hormonaalinen hoito (n=62).

Keskimääräinen seuranta-aika elinaika-analyysissä oli 77,7 kuukautta ja seuranta-ajan lopussa 84,4

% potilaista oli elossa. Elossaoloajan mediaani hormonaalista hoitoa saaneilla potilailla oli 53,8 kuukautta (95 %:n luottamusväli 44,5 – 63,2 kuukautta). Muissa hoitoryhmissä mediaania ei saavutettu seuranta-aikana. Eturauhassyöpä oli harvinainen kuolinsyy etenkin aktiivisen seurannan ja leikkaushoidon ryhmissä. Hormonaalisen hoitoryhmän potilailla oli alhaisin elämänlaatu ja elossaoloaika, ja näin ollen myös alhaisin määrä saavutettuja QALY:ja kahden vuoden aikana tutkimuksen alusta. Kolmen muun hoitoryhmien erot elämänlaadussa, kokonaiselossaoloajassa ja näin ollen myös saavutetuissa QALY:issa olivat vähäisiä. Tutkimus myös osoitti, että hoidon alussa joko 15D-lukemalla tai eräillä 15D-dimensioilla mitatuilla terveyteen liittyvällä elämänlaadulla on ennustearvoa arvioitaessa potilaiden hengissäpysymistä.

Kuten kirjallisuuskatsauksessa havaittiin, geneeristen, preferenssipohjaisten elämänlaatumittareiden käyttö eturauhassyöpäpotilailla on melko harvinaista. Tästä syystä tutkimuksessa tehtiin regressiomalli geneerisen 15D-lukeman ennustamiseksi sairaus-spesifillä EORTC QLQ-C30-datalla. Tämän “mapping”-mallin selitysaste oli 79 %, joka kertoo, että EORTC- instrumentin muuttujat selittivät hyvin 15D-lukemien vaihtelua.

(9)

Acknowledgements

This study was carried out in the Department of Public Health in the Faculty of Medicine in the University of Helsinki during 2009-2020. I wish to thank everybody involved in this work along the way. My special thanks belong to patients whose participation made this work possible.

I am deeply grateful to my supervisors, Professor Risto P. Roine and Professor Harri Sintonen, and thesis committee member Professor Kimmo Taari for your professional guidance and valuable advice. I feel privileged to have had this chance to work with you. You have always been available to discuss and guide this work and I am deeply grateful for the valuable time you have had to support me in this work.

I wish to express my gratitude to official reviewers of the thesis, Adjunct Professor Leena Forma and Professor Matti Laato, for their thorough review. Your valuable and constructive comments significantly improved this work. I am honored to have Adjunct Professor Juha Laine as the opponent in the public examination of the thesis.

I am truly thankful to all the people who have encouraged me both to start and to finish this work, especially Jarmo Hahl, Timo Muhonen, Niilo Färkkilä and Timo Seppälä. My warmest thanks to co- author Saku Torvinen, for inspiration, friendship and sharing the pain. I warmly thank all co-authors for their contribution in this work. I also express my deep gratitude to research nurses Merja Rignell and Paula Saari for their valuable work in the collection of the clinical data from the patient records.

Finally, my warmest thanks go to my family and friends. I am deeply thankful to my parents Marita and Jouko, and to my brothers Arttu and Perttu for your support and encouragement throughout the years. Special thanks to my digi-native little brother Perttu for IT-support and saving me from gray hairs. I thank all my friends for being there, in better or worse. With love I thank you, Tomi, for the happiness and love you bring to my life. And my precious children - Sofia and Sasu - the words cannot describe the amount of love I have for you. Every day you remind me of what is the most important in this life.

This work was financially supported with Pfizer Urologia scholarship which I acknowledge with gratitude.

Espoo 8.11.2020 Susanne Bergius

(10)

List of original publications

The dissertation is based on the following articles, which are referred to by their Roman numerals I, II and III.

I Torvinen S*, Bergius S*, Roine R, Lodenius L, Sintonen H, Taari K. Use of patient assessed health-related quality of life instruments in prostate cancer research: a systematic review of the literature 2002-15. International Journal of Technology Assessment in Health Care. 2016 Jan;32(3):97-106.

II Bergius S, Torvinen S, Muhonen T, Roine R, Sintonen H, Taari K. Health-related quality of life among prostate cancer patients: real-life situation at the beginning of treatment, Scandinavian Journal of Urology, 2017;51(1):13-19.

III Bergius S, Roine R, Taari K, Sintonen H. Health-Related Quality of Life and Survival in Prostate Cancer Patients in a Real-World Setting. Urologia Internationalis, 2020, DOI:

10.1159/000510319

*Authors share equal contribution

(11)

Abbreviations

ADT Androgen deprivation therapy

AQoL Assessment of quality of life

ANOVA Analysis of variance

CBA Cost-benefit analysis

CEA Cost-effectiveness analysis

CI Confidence interval

CMA Cost-minimization analysis

CRPC Castrate resistant prostate cancer

CUA Cost-utility analysis

EBM Evidence-based medicine

EBRT External beam radiation therapy

ECHOUTCOME European consortium in healthcare outcomes and cost-benefit research ECOG Eastern Cooperative Oncology Group

eHR Electronic health record

EMA European medicines agency

EORTC European Organization for Research and Treatment of Cancer EPIC Expanded prostate cancer index composite

FACT-G Functional assessment of cancer therapy - general

FACT-P PCS Functional assessment of cancer therapy - prostate cancer subscale

FDA Food and Drug Administration

GDP Gross domestic product

HAS Haute Autorite´ de Santé

HYKS Helsingin seudun yliopistollinen keskussairaala (Helsinki University Hospital)

HRQoL Health-related quality of life

HTA Health technology assessment

HTM Health technology management

HUI Health utilities index

HUS Helsingin ja Uudenmaan sairaanhoitopiiri (Helsinki and Uusimaa hospital district)

ICD International classification of diseases ICER Incremental cost-effectiveness ratio

ICHOM International Consortium for Health Outcomes Measurement ICUR Incremental cost-utility ratio

INAHTA International Network of Agencies for Health Technology Assessment IQWiG Institute for quality and efficacy in healthcare

LHRH Luteinizing-hormone-releasing hormone agonists

LRP Laparoscopic radical prostatectomy

MCDA Multi criteria decision analysis

MeSH Medical subject heading

mCRPC Metastatic castrate resistant prostate cancer

NHP Nottingham health profile

NHS National Health Service

NICE National Institute for Health and Care Excellence

OFT Office of Fair Trading

(12)

OLS Ordinary least squares

OS Overall survival

PC Prostate cancer

PFS Progression free survival

PRO Patient-reported outcome

PROM Patient-reported outcome measure

PSA Prostate-specific antigen

PC-QoL Prostate Cancer – Quality of Life PORPUS Patient Oriented Prostate Utility Scale

QALY Quality-adjusted life year

QoL Quality of life

QWB Quality of well-being scale

RARP Robotic-assisted radical prostatectomy

RCT Randomized controlled trial

RP Radical prostatectomy

RWD Real-world data

RWE Real-world evidence

SD Standard deviation

SF-36 Short form 36

SF-6D Short form six dimension

SG Standard gamble

SIGN Scottish Intercollegiate Guidelines Network

SMC Scottish Medicines Consortium

SRE Skeletal-related event

TNM Tumor Nodus Metastasis classification of malignant tumors

TTO Time trade-off

UCLA-PCI Expanded University of California Los Angeles prostate cancer index

VAS Visual analogue scale

VBHC Value-based health care

VBM Value-based medicine

VBP Value-based pricing

WHO World Health Organization

WTP Willingness-to-pay

ZIN Zorginstituut Nederland

Appendices

Appendix 1: 15D and EORTC-QLQ-C30 questionnaires Appendix 2: Published articles

(13)

1. Introduction

The ultimate goal of health care is to maximize health with given resources. The quantity of health can be expressed as quality-adjusted life years (QALY), which is a concept that takes into consideration both the length and the health-related quality of life (HRQoL). As resources given to healthcare are limited, treatment decisions must be made in the context of scarce resources. This is the context in which the discipline of health economics operates.

Health economics tries to answer questions of equal, fair, and efficient allocation of resources in health care. The development of medicinal and pharmaceutical technology provides new, sometimes expensive, treatment options at the same time as the aging of the population is setting increasing pressure on the management of health care budgets. Thus, we need to make resource allocation decisions that are based on objective scientific evidence regarding the effectiveness and costs of various treatment alternatives. Cost-effectiveness analyses (CEA) are one element of informed decision-making. However, CEAs do not provide unambiguous answers for decision- making, as ethical and moral questions must also be considered.

Health-related quality of life (HRQoL) tries to capture patients' physical, mental, and social domains of health, and it has importance both clinically as well as in health economics. In health economics, it is essential to be able to compare generic HRQoL between diseases and therapy areas for the purposes of cost-utility analyses evaluating the value of different health care interventions.

This study is done in the setting of real-life clinical practice, which sets specific conditions for the study. Measuring outcomes in real-life circumstances will be enhanced in the future by digitalization and, subsequently, improved data availability. Real-world data (RWD) will help to understand the impact of healthcare interventions in clinical practice and it is assumed that increased usage of RWD will also impact the decision making of the health care authorities (Eichler 2018, FDA 2019).

Effectiveness of healthcare can be measured in various ways by measuring both patients' subjective outcomes such as HRQoL as well as objective clinical parameters. The usability and validity of different outcome measures during different stages of a patients’ journey must be better understood to define their place in the overall assessment of the effectiveness of health care.

(16)

Generic HRQoL instruments may not be the most sensitive outcome measures in a given disease, but the importance of the utilization of these instruments arises from the need of a single and uniform outcome measure to appraise economic and clinical value of health care interventions across therapy areas. As QALY is chosen in some countries to be the uniform outcome measure used in cost-effectiveness analyses in Health Technology Assessments (HTA) of healthcare interventions and thus in prioritizing decisions (NICE 2013, Lääkkeiden hintalautakunta 2019), this creates also the framework for HRQoL measurement of prostate cancer (PC) patients in this study. As health economic evaluations based on QALY have health policy implications through prioritization decisions, the usability of QALYs in the context of economic evaluation and HTA is the focus of the background chapter of this study. In the background chapter I try to assess the place of QALYs in the overall value assessment framework of health care interventions. Considerations of the literature review about value assessment of interventions are not specifically linked to PC but are valid across disease areas.

One of the aims of this study was to collect HRQoL data and thus to understand the validity of a generic HRQoL instrument, the 15D, among prostate cancer (PC) patients in supporting the decision- making of treatment choices in PC. Findings of the empirical part of this study are specific for PC and thus not transferrable to other disease areas. PC as an area of interest arises from the fact that PC is becoming an increasingly prevalent disease due to aging population and improved survival of PC patients in Finland. Therefore, PC poses a significant burden to public health in Finland and involves many lives making clinical and economic implications of treatment choices significant for the health care, the patients and thus for the society.

(17)

2. Background and review of the literature

2.1. Prostate cancer

Prostate cancer (PC) is a carcinoma that develops in the prostate, which is a walnut-sized gland in the male reproductive system. The prostate gland is located in the pelvis, between penis and rectum, and is responsible for producing seminal fluid to mix with sperm from the testis, and thus to help sperm to travel and survive.

PC is the second most common cancer among men with 1.3 million new cases of prostate cancer and 359,000 associated deaths worldwide in 2018 (Bray et al. 2018). PC incidence rates are highest in industrialized western countries such as the Nordic countries, Western Europe and Northern America, but mortality rates do not follow those of incidence in these countries (Bray et al. 2018).

The introduction of prostate-specific antigen (PSA) testing for the detection of early-stage prostate cancers has been associated with a rapid increase in incidence rates, especially in higher-income western countries (Brawley 2012, Bray and Kiemeney 2017). There is variance in mortality rates at the global level (Bray et al. 2018), but in Europe and other Western countries, there has been a moderate decline in mortality rates (Bray and Kiemeney 2017). Figure 1 illustrates the incidence and mortality of PC in Europe.

(18)

Figure 1: PC incidence vs mortality in Europe based on age-standardized rates

Source: Bray and Kiemeney 2017

Relatively little is known about PC etiology, but age is the most important risk factor, and most of the cases are diagnosed in men over 70 years of age. Male hormones, ethnic and genetic background, and certain environmental factors are known risk factors. There is evidence that the importance of environmental factors is greater compared to inherited risk (Zaridze et al. 1984, Lichtenstein et al. 2000). Environmental risk factors include a diet containing high amounts of fat or meat and smoking (Hori et al. 2011, Zu and Giovannucci 2009). On the other hand, physical activity is known to reduce the risk of PC (Liu et al. 2011).

Early-stage PC usually does not cause symptoms, or they can be similar to those of benign prostatic hyperplasia. Early symptoms are mostly related to urinary dysfunction. Advanced prostate cancer can cause symptoms in sexual function, and metastatic disease that has spread to other parts of the body can cause additional symptoms such as bone pain.

(19)

2.1.1. Screening and diagnosis

PC mortality has decreased in the past decade, which is mainly attributed to the widespread use of PSA screening. While 1 in 7 men will be diagnosed with the disease during their lifetime, only one man in 30 with prostate cancer will die of the disease (Siegel et al. 2017). Still, PC remains the second leading cause of male cancer deaths. The dilemma in managing PC is the balance between the early detection of a potentially lethal disease that may benefit from treatment and the over-treating of low-risk (screening-detected) cancers that causes complications from the unnecessary treatment (Eastham 2017). This dilemma continues to be a controversy regarding prostate cancer screening and is one of the most debated topics in the urological literature (Loeb 2014).

Most of the PC cases are found through PSA testing. If PC is suspected, a tissue biopsy is needed for diagnosis. The Gleason grading is done using samples from the biopsy. Gleason grading is used for evaluating the prognosis of PC, and together with other parameters, it is used in PC disease staging, which helps to predict prognosis and guide therapy choices. The Gleason score is based on the microscopic appearance of PC cells. It ranges from 2 to 10, and PCs with a higher Gleason score are more aggressive and have a worse prognosis. The total Gleason score is a sum of two numbers. The first half of the score is based on the dominant cell morphology (scored 1—5), and the second half is based on the non-dominant cell pattern with the highest grade (scored 1—5). These two numbers are then combined to produce a total score of the PC (e.g. 3+3=6 for mild PC; 5+5=10 for very aggressive PC).

The identification and characterization of the disease have become increasingly precise in recent years. These advances are mainly due to the improved risk stratification, advances in magnetic resonance and functional imaging, as well as due to the emergence of several new biomarkers that can help to identify potential false-negative cases (Litwin & Tan 2017).

2.1.2. Treatments

There are multiple management options available for the treatment of PC and each treatment has its particular side effects. Most of the PC cases are detected at an early stage, and the treatment choice in localized PC is based on risk category (stage classification, histopathologic classification, and PSA), life expectancy of the patient (other morbidities, age, overall condition of the patient), expectations of the patient (adverse reactions from treatments, mental state, personality) and local

(20)

conditions (distances to treatments and treatment possibilities) (Prostate Cancer: Current Care Guidelines 2014). The treatment choice is made in discussion with the patient and possibly with his relatives so that the estimated life expectancy is put in proportion together with possible adverse reactions from treatments and risk of shorter life expectancy due to PC.

Active surveillance refers to a serial monitoring program for disease progression with the intent to take further action when needed. It appears to be a safe and more popular approach for men with less aggressive prostate cancer, particularly those with a PSA level of less than 10 ng/ml and tumors ǁŝƚŚ'ůĞĂƐŽŶƐĐŽƌĞϯരнരϯ;>ŝƚǁŝŶΘdĂŶϮϬϭϳͿ͘ /ŶĂĐtive surveillance, the aim is to avoid unnecessary treatment in men with localized PC who do not require immediate treatment, but at the same time to monitor the patient and aim at the correct timing for curative treatment if needed (Bruinsma et al. 2017). Patients are under close surveillance through a structured surveillance program with regular follow-ups. The need for curative treatment is prompted by predefined thresholds that are indicative of potentially life-threatening disease, which is still potentially curable.

Watchful waiting and active surveillance are sometimes confused with being the same thing, which they are not. Watchful waiting refers to conservative management of patients that are considered unsuitable for curative treatment right from the beginning. Patients are “watched” for the development of either local or systemic progression and when progression is detected patients are treated with palliative intention according to their symptoms in order to maintain quality of life (Mottet et al. 2019)

Radical prostatectomy (RP) is a surgery that targets eradication of cancer, while whenever possible, preserving continence and potency. Radical prostatectomy can be performed in an open, laparoscopic, or robot-assisted approach. There are a number of studies comparing different approaches for radical prostatectomy (Coughlin et al. 2018, Allan & Ilic 2016; Yaxley et al. 2016).

However, a Cochrane review comparing robotic-assisted radical prostatectomy (RARP) or laparoscopic radical prostatectomy (LRP) to open radical prostatectomy found no differences in oncological outcomes, urinary function or sexual function outcomes between the treatments.

However, RARP and LRP both resulted in statistically significantly shorter hospital stays and reduced need for blood transfusions over open RP (Ilic et al. 2017). No surgical approach, however, has so far been recommended over another by the European Association of Urology (Mottet et al. 2019).

(21)

Incontinence and sexual dysfunction are the main adverse effects of RP and can have a ƐŝŐŶŝĮĐĂŶƚ impact on HRQoL (O'Connor & Fitzpatrick 2006).

Major categories of radiotherapy (RT) in PC are external beam radiotherapy (EBRT) and brachytherapy (BT), which is an internal radiation. In EBRT, which includes multiple techniques of radiation, beams of radiation are focused on the prostate gland from outside of the body.

Brachytherapy uses small radioactive pellets, or “seeds,” which are placed directly into the prostate.

Brachytherapy alone is generally used only in men with early-stage PC. RT can result in early and late toxic effects. Additional side effects from radiotherapy can include bowel, urinary and erectile dysfunction. Modern radiation techniques have shown ƐŝŐŶŝĮĐĂŶƚ ƉƌŽŵŝƐĞŝŶ ĚĞůŝǀĞƌŝŶŐĞƐĐĂůĂƚĞĚ doses to the prostate without seriously affecting the side-ĞĨĨĞĐƚƉƌŽĮůĞ (O'Connor & Fitzpatrick 2006).

Hormonal therapy, also called androgen deprivation therapy (ADT), is a treatment that stops testosterone from being produced or reaching prostate cancer cells. As most PC cells need testosterone to grow, hormone therapy causes PC cells to die or grow more slowly. Testosterone- lowering therapy, i.e., castration, can be done by surgical or chemical castration. Long-acting luteinizing-hormone-releasing hormone agonists (LHRH) are currently the main form of ADT. Other traditional chemical ADTs are steroidal (such as. cyproterone acetate, megestrol acetate, and medroxyprogesterone acetate) or non-steroidal (such as nilutamide, flutamide, and bicalutamide) drugs. A problem with castration is that castration resistance (CRPC) will develop over time. This has led to the development of newer compounds for CRPC such as abiraterone acetate, enzalutamide, and apalutamide. Hormonal therapy is associated with a multitude of side effects that can impact HRQoL. Side effects from hormonal therapy can include osteoporosis, hŽƚ ŇƵƐŚĞƐ, fatigue, loss of energy, emotional distress, sexual dysfunction and metabolic syndrome (O'Connor & Fitzpatrick 2006, Iversen et al. 2000, Patil & Bernard 2018). In addition, hormonal therapy has been associated with neurocognitive deficits, thromboembolic disease, and cardiovascular disease, although the data regarding the associations are mixed (Patil & Bernard 2018).

The choice of therapy ultimately depends on its effectiveness. However, if there are no clear advantages of one therapy over another, the side-ĞĨĨĞĐƚƉƌŽĮůĞƐĂŶĚƚŚĞŝƌƌĞůĂƚŝǀĞŝŵƉŽƌƚĂŶĐĞƚŽ the patient remains in a crucial role (O'Connor & Fitzpatrick 2006). Findings from direct comparisons

(22)

in randomized clinical trials among patients with localized PC suggest that EBRT, BT and RP are all effective treatments for localized prostate cancer and that post-operative EBRT is also effective but might be associated with additional toxicity (Wolff et al 2015). Studies have shown that treatment for localized prostate cancer involving either EBRT, BT or RP can result in long-term erectile dysfunction (ED) (Alemozaffar et al. 2011, Ferrer et al. 2008). Also urinary and bowel dysfunction are common side effects of treatments in localized PC (Donovan et al. 2016, Barocas et al. 2017). In a study of men with localized PC, RP was associated with clinically significant declines in sexual function compared to EBRT and AS, especially in men with excellent function at baseline. Urinary incontinence function also declined significantly after surgery compared to EBRT and AS, with 14%

of RP patients reporting moderate or big problems with urinary leakage at 3 years, compared to 5%

with EBRT and 6% with AS (Barocas et al. 2017).

Chemotherapy has a relatively new role in PC, and it was only established in 2004 after demonstrating a survival benefit with docetaxel in metastatic castrate-resistant prostate cancer (mCRPC) (Boulos & Mazhar 2017). Docetaxel has been the most used chemotherapy with cabazitaxel as second-line therapy (Nader et al. 2018). In addition, there are new evolving biological, molecular-targeted therapies and immunotherapies. Also, new vaccines, hormonal therapeutics, and bone-targeting agents have demonstrated efficacy in men with metastatic prostate cancer resistant to traditional hormonal therapy (Litwin & Tan 2017).

2.1.3. PC in Finland

In 2017 in Finland, 5446 new PC patients were diagnosed and there were 912 PC deaths (Finnish Cancer Registry 2017). The relative 1-year survival from diagnosis was 98 % and the 5-year relative survival 93 % (Prostate Cancer: Current Care Guidelines 2014). The incidence of PC is increasing due to early diagnosis and aging of the population, but luckily, mortality rates are not following incidence rates due to good surveillance of early disease and advanced treatment options.

2.2. Value in healthcare

Value plays a vital role in health care systems. In the concept of value, both outcomes and costs are essential, but the definition of value is not precise and can be differently interpreted depending on the perspective. Patients, physicians, policy- and decision-makers can all have a different definition of value. However, as we are facing increasing budget pressure from the aging population, it has

(23)

become pronounced in the past decades that clinical value is not enough and things also have to be evaluated in terms of economic value. On top of traditional Evidence-Based Medicine (EBM) (Guyatt et al. 1992), the concept of value-based-medicine (VBM) (Campolina 2018, Bae 2015) or value-based healthcare (VBHC) (Tsevat & Moriates 2018, Porter 2010) has emerged. Researchers are proposing that in the concept of VBM, also patient empowerment and patient-centricity are key aspects (Marzorati & Pravettoni 2017). Measurement of HRQoL and quality-adjusted life years (QALYs) is a key concept in economic evaluations, but also other concepts to capture value in healthcare exist.

In the following chapters, I try to place the role of HRQoL into the perspective of overall value assessment in healthcare.

2.2.1. Outcome measures in PC

As PC is a heterogeneous disease, treatment options, and thus relevant outcome measures, vary depending on the disease stage. Some PC cases behave very aggressively, leading to metastases and eventually PC death, but most PCs, especially those detected early through PSA screening, have an indolent growth pattern. These PCs might never give rise to symptoms during a lifetime, even if left untreated. Radical treatment with surgery or radiotherapy of such cancers would thus result in undesirable side effects and deterioration of quality of life (Litwin & Tan 2017). The success of care in such cases is the combination of the absence of disease progression and good HRQoL. For localized PC, evolving patterns of care, including the increased use of active surveillance and the development of novel therapeutic options, are expected to have positive effects on HRQoL.

Consequently, relevant outcome measures for follow-up require the use of patient-reported outcome measures (PROMs or PROs) such as HRQoL. Researchers point out that also in the assessment of long-term impacts of the disease, the use of PROs is required (Connell et al. 2019). It has been recommended already more than 20 years ago that to understand all aspects of outcomes in PC, both generic HRQoL and disease-specific measures must be used (Litwin et al. 1995).

In advanced PC, disease-progression is usually monitored more closely than in localized disease. In trials of advanced PC, the most common primary outcomes are overall survival (OS) and progression-free survival (PFS). A review assessing clinical trial endpoints in advanced PC did not find any trials using a QoL or pain outcome as a primary or co-primary endpoint (Fabricius et al.

2015). Another review found that in trials of metastatic castration-resistant PC patients, PROs were either not being measured routinely, or if used, were often not reported adequately (Fallowfield et

(24)

al. 2016). It has been emphasized that more focus should be placed on assessing PROs also in later stages of the disease (Morgans & Stockler 2019).

The International Consortium for Health Outcomes Measurement (ICHOM) has published guidelines about which patient-derived outcome measures would be useful to collect in different stages of PC.

For men with early PC, ICHOM recommends that the disease-specific EPIC-26 and Utilization of Sexual Medications/Devices are collected as PROs. Other outcome measures to be collected are survival and disease progression parameters such as biochemical recurrence (ICHOM 2017).

Moreover, for patients with advanced PC, Eastern Cooperative Oncology Group (ECOG) or WHO performance status, pain medication usage, EORTC QLQ-C30 questionnaire (for pain, fatigue, physical and emotional functioning) and additional parameters for disease progression such as symptoms from skeletal-related-event (SRE) are advised to be collected (ICHOM 2017).

2.2.2. Health-related Quality of Life

HRQoL is one type of PROMs. It is a broad concept reflecting a person’s functioning in life and how a person perceives his/her health. The World Health Organization (WHO) defines health as a state of physical, mental, and social well-being, and HRQoL is functioning and well-being in relation to health. The history of measuring health status can be traced to the early 1970s, and measuring was motivated by a desire to measure outputs and performance of health care systems (Fanshel & Bush 1970).

There are both generic instruments that provide a summary of HRQoL, and disease-specific instruments that focus on problems associated with single disease states, patient groups, or areas of function. Disease-specific instruments are designed for assessing health in particular conditions, and they are not suited for comparison across interventions or populations. Therefore, to compare health between various diseases, interventions, or populations, generic HRQoL instruments should be used. Generic instruments can further be categorized into those providing health profiles and preference-based measures that generate health utilities, usually values between 0 and 1 (Guyatt et al. 1993), which allows comparison of cost-utility in health economic evaluations.

(25)

2.2.2.1. Prostate-specific HRQoL instruments

Common disease-specific PROs specifically used in PC are listed in table 1. The FACT-P PCS and EORTC QLQ-PR25 are PC-specific modules of the cancer-related QoL instruments FACT-G and EORTC QLQ-C30, respectively. The EORTC QLQ-PR25, FACT-P, and PORPUS are designed for all tumor stages, whereas EPIC, PC-QoL, and UCLA-PCI are specifically designed for patients at an early stage of the disease. Instruments are different with respect to health domains they include, and EORTC- QLQ-PR25 and EPIC are the only instruments that take into account the whole spectrum of symptoms in the urinary, bowel, sexual and hormonal domains (Schmidt et al. 2014).

Table 1: Prostate-specific HRQoL instruments

Disease-specific instrument Abbreviation Tumor stage* Source Expanded Prostate Cancer Index

Composite-26 EPIC early stage Wei et al. 2000

Expanded University of California-Los

Angeles Prostate Cancer Index UCLA-PCI early stage Litwin et al. 1998 Functional Assessment of Cancer

Therapy - Prostate Cancer Subscale FACT-P PCS all Cella et al. 1993 European Organization for Research

and Treatment of Cancer Quality of Life Questionnaire prostate specific

EORTC QLQ-PR25 all van Andel et al. 2008 Prostate Cancer – Quality of Life PC-QoL early stage Giesler et al. 2000 Patient Oriented Prostate Utility

Scale PORPUS all Krahn et al. 2000

*recommended tumor stage

Many of the prostate cancer-specific PROs claim that the instruments measure HRQoL or overall QoL, but their dimensions focus on urinary, sexual and bowel symptoms, and functioning. Their main focus is thus on the physical impact of the disease, and less attention is paid to the mental or social dimensions. A review assessing the usefulness of PROs among PC patients undergoing radical surgery concluded that there are gaps in their content and inadequate evidence of reliability, validity, and responsiveness, as well as their suitability for use in clinical practice with individual patients (Protopapa et al. 2017). Nevertheless, researchers recommend that HRQoL should be used more widely both in clinical trials as well as to inform patients and regulatory agencies on HRQoL aspects of therapies (Morgans & Stockler 2019).

(26)

Enhanced tumor detection with PSA testing has moved PC diagnoses to younger patients at earlier stages, and men are living longer with the knowledge of the disease and possible side effects of treatments. Disease-specific instruments have an important role in the evaluation of benefits and harms to PC patients. The responsiveness, i.e., ability to detect the change when it has occurred, may be better in some disease-specific instruments than generic HRQoL instruments, and the PORPUS questionnaire was found to be more sensitive than certain generic instruments (Krahn et al. 2007). The most obvious explanation for this is that the health domains of greatest importance in HRQoL following prostate cancer diagnosis and early treatments are often sexual, urinary, and bowel function. None of the generic instruments in the study by Krahn et al. (2007) included any items related to sexual function.

2.2.2.2. Generic HRQoL instruments

An advantage of generic HRQoL instruments is that they are applicable across a wide range of populations and thus allow comparison of HRQoL between different diseases and therapy areas.

Generic health profile instruments include the widely used SF-36 (Stewart et al. 1992, McEwen &

McKenna 1996) and the Nottingham Health Profile (NHP). Health profile instruments provide multiple outcome scores that can be useful to clinicians and/or researchers when attempting to measure differential effects of conditions or treatments on various HRQoL domains. However, they do not produce a single index score needed for cost-utility analyses. Consequently, the SF-36 has been revised into a six-dimensional health state classification called the SF-6D, which is a preference-based measure providing a single index score for economic evaluations (Brazier et al.

2002).

Similarly, also other preference-based HRQoL measures provide a single number, usually between the continuum from perfect health (1) to death (0), although other scales also exist. The health index score represents the respondent’s subjective health status and incorporates a preference value (utility) for that overall health state. Utilities can be elicited in two different ways: either by direct or indirect valuation methods (Brazier et al. 1999).

Valuation methods include such approaches as the Standard Gamble (SG) (Torrence 1976), the Time-Trade-Off (TTO) (Dolan et al. 1996), the Rating Scale (RS) (Rosser and Kind 1978), and the Visual Analogue Scale (VAS) (Gudex et al. 1996). In the TTO method, respondents’ preferences are

(27)

examined by asking what they value equally - living in a given health state for a certain period of time, or a shorter time in full health. In the SG method, respondents are choosing between a certain outcome in a given health state or a gamble with a probability (p) for the best possible outcome and a probability (1-p) for the worst possible outcome, usually dying. In the VAS method, respondents are asked to rate their health state on a continuous rating scale e.g., from 0 (worst possible health, dead) to 100 (perfect health). An advantage of the VAS method is its simplicity (Brazier et al. 1999).

In the direct valuation people either value their own health or the health states to be valued are described in a written form in their entirety to those, from whom the valuations are elicited (usually members of the general public), and they must imagine themselves in these hypothetical states even if different valuation methods are used. In the indirect approach, a small set of health states is valued directly and using these data, values for a wider set of health states are predicted by regression techniques. Or health states are split into parts and these parts are then valued separately and finally aggregated to values of different health states. Then, different health states defined by generic HRQoL questionnaires are weighted with these values or preferences to represent the values of the community regarding the appreciation of different health states (Brazier et al. 1999). The most commonly used multi-attribute utility instruments are introduced in table 2.

These instruments provide a framework for respondents to describe their health states, to which preference values are then applied from population-based preference functions to calculate a single index utility score.

Table 2: Generic multi-attribute, single index HRQoL instruments Generic instrument Abbreviation Source

EuroQol EQ-5D Brooks 1996

Health Utilities Index, Mark II/Mark III HUI Torrance et al. 1982

Short Form 6D SF-6D Brazier et al. 2002

Assessment of Quality-of-Life AQoL Hawthorne and Richardson 2001

15D 15D Sintonen 1981

2.2.2.3. Quality-adjusted life years

A commonly used application of utility is quality-adjusted life years (QALY), which combines both the quality and length of life. The idea of calculating QALYs is straightforward - the amount of time

(28)

spent in a given health state is weighted by the utility score given to that health state. Thus, one year of perfect health (utility score of 1) generates one QALY, and two years in a health state valued with 0.5 is also worth one QALY. As utilities, the number of QALY gained provides a common currency to assess the extent of benefit gained by different healthcare interventions in terms of HRQoL and survival and also allows comparisons between interventions. When the number of QALYs gained is combined with the costs associated with interventions, they provide an assessment of the relative value of the intervention, i.e., the worth of the intervention from an economic perspective. The number of QALYs gained combined with costs incurred generates a comparable cost-utility ratio. Figure 2 illustrates a situation in which intervention B provides a better utility compared to intervention A through the time. The difference in QALYs can be calculated as the difference in the areas under the curves for interventions A and B (Drummond et al. 2005).

Figure 2. QALYs gained between intervention A and B

The concept of QALYs is not without critique. The QALY approach has been criticized on technical and ethical grounds (Prieto & Sacristán 2003, Sanders et al. 2016). One of the technical issues has to do with the choice of utility instruments, which are known to provide different results and thus impact the cost per QALY comparison (Whitehurst et al. 2014). There is no consensus of a gold standard regarding the most appropriate generic preference-based measure of utility. Other areas of controversy include the limitation of the QALY approach in terms of the health benefits it can capture, its blindness towards equity concerns, and the underlying theoretical assumptions. A growing debate is related to whether a QALY is the same regardless of to whom it accrues and also

(29)

to the issue as to who should value health states (Whitehead & Ali 2010). The European Commission project “European Consortium in Healthcare Outcomes and Cost-Benefit Research (ECHOUTCOME)”

was studying how 27 European health system organizations use health outcomes in the frame of Health Technology Assessments (HTA) and the robustness of the QALY as an indicator of health. The recommendation of the project was that QALY assessment for health decision-making should be abandoned due to limitations and controversies of the QALY approach and cost-effectiveness analyses should rather be expressed as costs per relevant clinical outcome (Beresniak et al. 2015).

However, QALY assessments are still central in decision-making in Europe, and no other approach has so far proved to be more robust. Germany adopted an ”efficiency frontier” approach to compare the efficiency of new technologies to existing ones within disease classes using disease-specific metrics, rather than cost per QALYs approach for cross-disease comparisons (Caro et al. 2010). This approach could result in inequities, and political tension as different cost-effectiveness thresholds might be used for different therapeutic areas. On the other hand, the same issue can still exist in the cost per QALY approach as the pure cost/QALY ratio is usually not, and also should not be, the only criterion for decision-making.

The use of QALYs in cost-utility analyses has been the approach in HTAs in Europe but not traditionally in the United States (Neumann & Greenberg 2009). The need to deliver health care efficiently, and the importance of using analytical techniques to understand the clinical and economic consequence of interventions, has increased and also in the US there is a relatively recent recommendation to use the cost/QALY approach in HTAs with the understanding that it cannot be the mere basis for decision-making (Carias et al. 2018, Sanders et al. 2016).

2.2.3. Health economic evaluation

Health economic evaluations are needed to understand the relationship between health outcomes and investment needed, i.e., what is the worth of a health care intervention. Economic evaluation requires systematic identification, measurement, and valuation of inputs and outcomes of comparative technologies at issue (Drummond et al. 2005). Economic evaluations are most commonly employed in the context of health technology assessment (HTA) when new medicines and other technologies are introduced to healthcare systems. In Finland, there are guidelines on how to perform health economic evaluation for medicines (Lääkkeiden hintalautakunta 2019), but similar guidelines for other technologies do not exist. In the UK, there are guidelines to do

(30)

appraisals, not just for medicines, but also for medical devices, diagnostic techniques, surgical procedures, and health promotion activities (NICE 2013).

The most commonly used methods for health economic evaluation are cost-minimization analysis (CMA), cost-benefit analysis (CBA), cost-effectiveness analysis (CEA) and cost-utility analysis (CUA).

CMA assumes equal effectiveness of comparative technologies, which then allows a simple comparison of costs, and the logical decision is to choose the least expensive option. In CBA, also outcomes are expressed in monetary terms in order to calculate a net benefit. In CEA, outcomes are measured using “natural units”, such as events avoided, change in cholesterol level, or hospital days avoided. CUA is a specific case of CEA in which outcomes are expressed as QALYs gained (Drummond et al. 2005).

Table 3: Types of economic evaluation methods

Method Cost Outcome Expression of cost per

outcome Cost-minimization

analysis (CMA) Monetary Equivalence of outcomes in comparative treatments

Difference in costs of comparative treatments Cost-benefit analysis

(CBA) Monetary Monetary Net benefit = outcomes -

costs Cost-effectiveness

analysis (CEA) Monetary Single "natural" unit cost per outcome measure e.g. cost/avoided event Cost-utility analysis

(CUA) Monetary QALY cost/QALY gained

Adopted from Drummond et al. (2005).

2.2.3.1. Incremental cost-effectiveness

Decision-making about health care resource allocation can be complex, often requiring decision- makers to consider trade-offs, values of the patients and the society, and other types of evidence in the face of uncertainty and affordability. However, decision-making based purely on cost- effectiveness is fairly simple, and one needs only to decide which treatment is the better option. A cost-effectiveness plane was introduced to health care as an aid for decision-making in different situations (Black 1990) (Figure 3). If treatment is both less costly, and it provides better outcomes, it is a strongly dominant option (quadrant II in Figure 3) and should be chosen. The decision in quadrant IV is also clear and should not be chosen as it is both more expensive and less effective.

(31)

Decisions in quadrant I and III should be made based on incremental cost-effectiveness, i.e., the comparison of the difference in costs over the difference in outcomes to derive incremental cost- effectiveness ratio (ICER):

ܫܥܧܴ= ܥ݋ݏݐ (ܣ)െ ܥ݋ݏݐ (ܤ)

ܱݑݐܿ݋݉݁ (ܣ)െ ܱݑݐܿ݋݉݁ (ܤ)

In the case of extended (weak) dominance, there is a combination of two treatments that shows greater cost-effectiveness than a third one. Thus, in extended dominance, an intervention that has an incremental cost-effectiveness ratio that is greater than that of a more effective intervention is ruled out (Drummond et al. 2005).

Figure 3: Incremental cost-effectiveness plane, intervention vs. comparator (C)

Source: Adapted from Drummond et al. 2005 2.2.3.2. Willingness-to-pay

In practice, the situation is often as illustrated in quadrant I in figure 3, in which a new treatment is both more costly and more effective than the comparative treatment. In those cases, decisions must be made based on willingness-to-pay (WTP) for additional effectiveness. WTP is thus a maximum price at or below which the society (or a consumer) will buy a product or service (Varian 1992), or

(32)

in the case of CUA one QALY. The WTP threshold eventually defines if an intervention is cost- effective or not. Visual interpretation of the WTP threshold line in figure 3 is that treatments below the threshold line are considered cost-effective and those above not.

Not many countries have an explicit threshold for maximum WTP for a QALY, but those countries that do, mainly fall within the WHO’s recommended range of one-to-three times gross-domestic- product (GDP) per capita (Cameron et al. 2018). For example, in Finland, politicians or health authorities have not explicitly stated any range of an acceptable cost/QALY ratio. A traditionally referenced value in American health economic literature is the value of 50 000$ USD per QALY, which may arise from the cost of dialysis in the 1980s but lacks any scientific justification (Neumann et al. 2014). In the US, interventions in the cost/QALY range of $50 000-$100 000 are often reported to be cost-effective (Shiroiwa et al. 2010). In the UK, the National Institute for Health and Care Excellence (NICE) recommends a value of 20 000 - 30 000 UK pounds per QALY, which represents an informed estimate of the health forgone, based on the evidence that is available about the productivity of other NHS activities (Culyer et al. 2007).

2.2.3.3. Perspective

The perspective of an economic evaluation depicts the point of view that is adopted when deciding which types of costs and health benefits are to be included in an economic evaluation. The perspective taken is an important element and has an impact on the analysis. Typical viewpoints are those of the patient, hospital/clinic, other providers, healthcare system, or society. The broadest perspective is societal, which reflects a full range of social opportunity costs. The International Society for Pharmacoeconomics and Outcomes Research (ISPOR) Task Force defines the full societal perspective to include three conditions: 1) the inclusion of time costs, 2) the use of opportunity costs, and 3) the use of community preferences, which in practice very rarely takes place (Garrison et al. 2010). A typical approach is to include productivity losses arising from patients’ inability to work but the full societal perspective includes also relevant non–health-related impacts in other sectors such as in education and legal aspects, thus it is understandable that the full societal perspective is rarely taken (Garrison et al. 2018, Drost et al. 2017). It has been proposed that the terms “restricted” or “limited” societal perspective, defined as analyses including indirect costs and using community preferences, should be used as other types of analyses are often too theoretical.

(33)

It has also been emphasized that one should be explicit when using the healthcare system perspective or the payer perspective compared to a true societal perspective (Garrison et al. 2018).

A typical perspective adopted by HTA agencies, e.g., by the NICE in the UK, is the perspective of a healthcare system or provider, recognizing that the societal perspective may bias against those not working, such as retired persons or those not able to work due to health reasons. Thus, costs not to be included are patients’ costs of obtaining care such as transportation, over-the-counter purchases, co-payments, or time off work (NICE 2013).

The use of the term “societal perspective” is often adopted merely based on the choice of including productivity costs. In a systematic review assessing which costs were included in economic evaluations that were said to adopt the societal perspective, only a few studies included in addition to productivity costs also other costs such as informal or social care costs (Drost et al. 2017).

Measuring and interpretation of QALYs have been argued to be problematic depending on the perspective that economic evaluation adopts. It has been argued that including indirect costs to the perspective of analysis involves double counting if those effects are considered in the QALY measure. However, there is evidence indicating that productivity costs due to morbidity are not captured within individuals' health state valuations (Davidson & Levin 2008). These findings, therefore, suggest that productivity costs due to morbidity should be included as a cost in cost- effectiveness analyses. For QALYs to be interpreted as only a measure of health benefit, productivity effects need to be explicitly excluded when the value of a health state is assessed (Jönsson 2009).

However, as QALYs are often based on the general population’s valuation of health outcomes, in that sense the QALY is capturing a societal perspective.

2.2.4. Health Technology Assessment

Health Technology Assessment (HTA) is originally defined by the International Network of Agencies for Health Technology Assessment (INAHTA) as “a multidisciplinary field of policy analysis, studying the medical, economic, social and ethical implications of development, diffusion and use of health technology” (Luce et al. 2010). HTA is used to assist decisions about reimbursement and funding of new technologies. Economic evaluation forms a core element of an HTA assessment guiding decisions on resource allocation. However, the ICER approach lacks the elements needed for

(34)

decision-making, such as ethics and equity (Saarni et al. 2008). In addition, it lacks elements such as affordability, budget impact, or feasibility of implementation of an intervention, which are also crucial elements needed for decision-making. Criteria related to the disease (such as severity of disease, capacity to benefit, and past health loss), criteria related to characteristics of social groups an intervention targets (socioeconomic status, area of living, gender, race, ethnicity, religion and sexual orientation), and non-health consequences of an intervention (financial protection, economic productivity, and care for others) have been proposed to be included in the decision- making (Norheim et al. 2014). While HTA, in general, has a societal policy perspective, many agencies in practice take a narrower healthcare or provider budget perspective when performing economic evaluations (Jönsson 2009).

Even though other criteria, such as severity and equity, can be assessed along with ICERs in a full spectrum HTA, there are concerns that such approaches may fail to capture other important sources of value if they are not quantifiable in cost-effectiveness analysis. Baeten and colleagues (2010) have tried to capture equity into a quantifiable element to be included in cost-effectiveness analysis (Baeten et al. 2010), but these types of analyses are scarce.

Value dimensions often outside the scope of an ICER can include the value of innovation to the society, unmet need, disease severity, reduced caregiver burden and patient compliance (Goldman et al. 2010). Some of these dimensions could be quantifiable in ICER assessment, but some elements are in practice difficult to capture in a reliable and systematic way. If value dimensions are not quantifiable in an ICER, Goldman and colleagues (2010) suggest a two-part protocol in which technologies are scored along key dimensions of value neglected by most current HTA approaches.

Value dimensions would be scored publicly and transparently, and a composite ‘value score’ would be constructed. The composite value score could be considered jointly with the ICER (Goldman et al. 2010).

2.2.4.1. Multi-criteria Decision Analysis

The approach Goldman et al. (2010) proposes is similar to the Multi-criteria Decision Analysis (MCDA) approach that has also been suggested as a method to capture the benefits beyond QALYs in a transparent and consistent manner (Thokala & Duenas 2012). Multi-criteria decision analysis

(35)

frameworks have been suggested to offer a more holistic perspective to value assessment compared to traditional HTA (Angelis & Kanavos 2016).

MCDA has been defined as “an extension of decision theory that covers any decision with multiple objectives. It has been characterized as a methodology for appraising alternatives on individual, often conflicting criteria, and combining them into one overall appraisal...” (Keeney & Raiffa 1993).

Another definition by Belton and Stewart is that MCDA is “an umbrella term to describe a collection of formal approaches, which seek to take explicit account of multiple criteria in helping individuals or groups explore decisions that matter” (Belton & Stewart 2002).

There are areas of uncertainty involved with using MCDA in the HTA process, such as uncertainty of problem structuring, evidence, and variation in preferences (i.e., uncertainty in performance scores, criteria weights, thresholds, etc.) (Thokala & Duenas 2012). These uncertainties are similar to those related to traditional cost-effectiveness modelling. A systematic review reported that there is currently interest in MCDA in healthcare, which is mirrored in an increase in the application of MCDA to evaluate healthcare interventions. However, there are many MCDA methods available, which can be challenging (Marsh et al. 2014). The ISPOR task force has established a common definition for MCDA in health care decision making and developed good practice guidelines for conducting MCDA to aid health care decision making (Thokala et al. 2016). The use of MCDA in health care is in early stages; thus, good practice guidelines can only be considered “emerging” (Marsh et al. 2016).

2.2.5. Value-based healthcare

Value-based healthcare (VBHC) is a broad concept without just one unambiguous definition. An expert group set up by the European Commission defined VBHC in a recent opinion paper as actions needed to ensure the financial sustainability of universal healthcare and a reallocation of resources from low to high-value care. The same expert group stated, that as consistent and common language or practices about VBHC are not yet in place, much work and investments are needed in piloting, monitoring, and evaluating the reallocation and shifting of resources (EXPH 2019). Another definition of value-based healthcare is that by Porter and Teisberg (2006), who define value in health care as health outcome per money expended. This is similar concept to ICER, but outcomes obtained are divided by the costs. What Porter brings on top of traditional health economic theory are the elements of competition and dynamic improvement of effectiveness in health care. Porter

Health-related quality of life and survival outcomes in prostate cancer

HEALTH-RELATED QUALITY OF LIFE AND SURVIVAL OUTCOMES IN PROSTATE CANCER

SUSANNE BERGIUS

HEALTH-RELATED QUALITY OF LIFE AND SURVIVAL OUTCOMES IN PROSTATE CANCER

Susanne Bergius

Abstract

Tiivistelmä

Acknowledgements

List of original publications

Abbreviations

Appendices

CONTENTS

1. Introduction

2. Background and review of the literature